)]}' { "log": [ { "commit": "1195734e94f3842db7f9044d581d88494a8440db", "tree": "a404306592d5647fa71aa6cf817d4cc3f42169c2", "parents": [ "15eee605512951b018939dd2cf096722f2be4807" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 09 13:58:41 2024 +0100" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Jan 07 15:09:46 2025 +0000" }, "message": "treewide: remove toolchains_protoc\n\nAs discussed we want to reduce external binaries and therefore don\u0027t\nneed this.\n\nChange-Id: Ia3d053a9584baf54b6aefe79b88ac31126d54e14\nReviewed-on: https://review.monogon.dev/c/monogon/+/3687\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "15eee605512951b018939dd2cf096722f2be4807", "tree": "0f410396bd7bd7a5cfc5624184bcb5e791d23d80", "parents": [ "ccabae947fe90a36055e7a16f42c030a1c89f1ac" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Thu Jan 02 13:31:31 2025 +0100" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Jan 06 21:10:14 2025 +0000" }, "message": "m/n/k/containerd: update to 2.0.1\n\nPull in the first containerd 2.0 patch release as it fixes a bunch of\nuserns-related issues (like pipe ownership).\n\nChange-Id: I371f48fa321d7ed18f5e6b480bc27f9f1592a8a3\nReviewed-on: https://review.monogon.dev/c/monogon/+/3738\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "ccabae947fe90a36055e7a16f42c030a1c89f1ac", "tree": "6d292c851b180fbe1aa2ac27cbf788992e0f7e15", "parents": [ "35cd44be9b5f43d19e05f48f6be2ab6cd65b9c22" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Dec 17 22:25:33 2024 +0100" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Jan 06 15:36:47 2025 +0000" }, "message": "build/bazel: update chrony source url\n\nThe old one is not reachable anymore\n\nChange-Id: I166238408f3f330f87971af5c1f06e7a259ae89d\nReviewed-on: https://review.monogon.dev/c/monogon/+/3741\nReviewed-by: Leopold Schabel \u003cleo@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "35cd44be9b5f43d19e05f48f6be2ab6cd65b9c22", "tree": "1ec0672614bb0e03246dd91f03c3113a6f42c95a", "parents": [ "117006e5a66f0c5858bff44aa05dc036083610a3" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 16 02:44:11 2024 +0100" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Fri Jan 03 15:40:37 2025 +0000" }, "message": "osbase/supervisor: allow access to LogTree from Runnable\n\nThis adds a way to access the underlying LogTree instance from\na runnable with the same API as normal logger access. This is required\nfor migrating metropolis/node/core to use bringup.\n\nChange-Id: I867ff0ff104bd7f524667a66aaaf13fc29c8f25c\nReviewed-on: https://review.monogon.dev/c/monogon/+/3704\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "117006e5a66f0c5858bff44aa05dc036083610a3", "tree": "62491f59c1747dca7c2334c32dd1d22e6074195b", "parents": [ "8eeae7b7fb8f98497caa5c69f9d6528fbdd81281" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 09 22:53:52 2024 +0100" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Fri Jan 03 15:40:32 2025 +0000" }, "message": "third_party/linux: set transition overrides to default value\n\nWithout this a run with the race detector doesn\u0027t work as we prevent CGO\nwith this transition.\n\nChange-Id: Ibba17ec4058ed34bee1beb7b5f3c6cac345bac5d\nReviewed-on: https://review.monogon.dev/c/monogon/+/3690\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "8eeae7b7fb8f98497caa5c69f9d6528fbdd81281", "tree": "ca22c3af862b7f6641922b9878d3632c1fd561e4", "parents": [ "5241968495198eb1957b893071eac3d2bb7ca1b4" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 09 22:52:55 2024 +0100" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Fri Jan 03 15:40:32 2025 +0000" }, "message": "cloud/shepherd/manager: fix data race\n\nChange-Id: I74dfe7ebd274a2014df827ae804d02cded3af090\nReviewed-on: https://review.monogon.dev/c/monogon/+/3689\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "5241968495198eb1957b893071eac3d2bb7ca1b4", "tree": "bebab9899d5858d72e60f22f945be3cd06627a26", "parents": [ "837cb8e459b9eefabe89ab17df0b7dafb5e3d631" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Dec 23 15:57:35 2024 +0100" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Dec 23 21:59:59 2024 +0000" }, "message": "gomod: remove x/exp replace directive\n\nUpstream has since fixed the mentioned dependency problem, thus we can\nuse the normally-resolved version again.\n\nChange-Id: Ic9e7e0473d62c5adf89b1c0f9804296074dcd966\nReviewed-on: https://review.monogon.dev/c/monogon/+/3723\nTested-by: Jenkins CI\nReviewed-by: Jan Schär \u003cjan@monogon.tech\u003e\n" }, { "commit": "837cb8e459b9eefabe89ab17df0b7dafb5e3d631", "tree": "32337d84d4f32b0c2c523e2c5bd177f4acfe4808", "parents": [ "b6afed68fd1d2ee9b32d395b388d2db1338d0fa0" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Dec 23 13:52:56 2024 +0100" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Dec 23 21:59:59 2024 +0000" }, "message": "treewide: update Kubernetes to 1.32\n\nRelatively easy change, one cadvisor fix is temporarily needed. The\nlegacy log dir patch needed to be rebased, that\u0027s about it.\n\nI enabled single-process OOM killing again as that was the default for\ncgroupv1 and IMO the more sane behavior.\n\nUpstrem changelog at:\nhttps://github.com/kubernetes/kubernetes/blob/master/CHANGELOG/CHANGELOG-1.32.md\n\nChange-Id: I537a6e37137d05efb6eec8635915e36fd8b37cbc\nReviewed-on: https://review.monogon.dev/c/monogon/+/3721\nReviewed-by: Jan Schär \u003cjan@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "b6afed68fd1d2ee9b32d395b388d2db1338d0fa0", "tree": "9d4deac20acc468c78be9f55574c08035b9b26fc", "parents": [ "139550a8c8ff4c0d0c7557b9e5af0c76ea09a4ef" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Dec 23 13:04:34 2024 +0100" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Dec 23 21:59:59 2024 +0000" }, "message": "third_party/gvisor: remove old cgroups import\n\nWe don\u0027t want to use the old cgroups/v1 import, remove it.\nMost code using it is no longer in use as it has been replaced by\ncontainerd shim code, so just delete that.\nAdjust the rest to work with cgroups/v3.\n\nAs K8s 1.32 drops its dependency on cgroups/v1 Gazelle mangles the deps\nif we leave it in.\n\nChange-Id: Ic95fce7a17acbddf70a1b5b2a0978b9246a17c58\nReviewed-on: https://review.monogon.dev/c/monogon/+/3720\nTested-by: Jenkins CI\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\n" }, { "commit": "139550a8c8ff4c0d0c7557b9e5af0c76ea09a4ef", "tree": "88132557266dc11bcc65b85d7a1b3dc6ce741b9b", "parents": [ "a5b00bd43cceaf807e56ae609c6acdb9f94fa983" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Tue Jun 20 13:30:27 2023 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 23 13:11:49 2024 +0000" }, "message": "build/platforms: use our platform as host platform\n\nChange-Id: I5e5b1a6c629e97a034474ad36ff1d025f7bd4873\nReviewed-on: https://review.monogon.dev/c/monogon/+/1833\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "a5b00bd43cceaf807e56ae609c6acdb9f94fa983", "tree": "5c2c0fe0d3947ddc7d82f6c4b34bedaeb26b7f39", "parents": [ "70e5a977302e5ce90c94488b7da0d55894dea3ec" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 09 22:52:31 2024 +0100" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 23 13:11:29 2024 +0000" }, "message": "metropolis/test/launch: fix data race\n\nChange-Id: Id427d73cf32ff06a074fce5903d66050dd9e28cb\nReviewed-on: https://review.monogon.dev/c/monogon/+/3688\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "70e5a977302e5ce90c94488b7da0d55894dea3ec", "tree": "fc21c3671818973eb60f5e31ff753a86592f7ff0", "parents": [ "5b38443b6ef6337db8d4ed12d7d2dc59c0a079da" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Tue Jun 20 13:19:20 2023 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 23 13:11:29 2024 +0000" }, "message": "build/toolchain/musl-host-gcc: add pie feature, disabled by default\n\nThis isn\u0027t enabled anywhere, but we should enable it for all of our\nbinaries.\n\nChange-Id: I77dc163a7085c792c0b0d64e6423e9117b596ec4\nReviewed-on: https://review.monogon.dev/c/monogon/+/1838\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "5b38443b6ef6337db8d4ed12d7d2dc59c0a079da", "tree": "b0a97ea2c180aec385b71bbf8c8cd4185d927943", "parents": [ "3c6183f7f12ded0c563239c7eff1f1dc4a9bebec" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Tue Jun 20 13:17:30 2023 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 23 13:11:29 2024 +0000" }, "message": "third_party/musl: build PIE/PIC static library\n\nThis allows us to then build static-pie binaries.\n\nChange-Id: Ied5c34da85608df0c1b56aec321b9eec367a9615\nReviewed-on: https://review.monogon.dev/c/monogon/+/1837\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "3c6183f7f12ded0c563239c7eff1f1dc4a9bebec", "tree": "e3487e97b71fb10a310780a36c7438f4fb9af242", "parents": [ "bc32bcd241a9270786a749266ce4f6a5df6d0a6e" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 16 02:42:21 2024 +0100" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 23 13:10:21 2024 +0000" }, "message": "osbase/supervisor: use MustRegister for metrics registration\n\nChange-Id: I4321c626f210bea025ab27bfecf783425f1482b5\nReviewed-on: https://review.monogon.dev/c/monogon/+/3703\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "bc32bcd241a9270786a749266ce4f6a5df6d0a6e", "tree": "d82d87a2af96a2cf40894869de5550de8500560e", "parents": [ "5b997a113d011c35ecc981594c95a53d67311f83" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Dec 18 17:36:41 2024 +0100" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Dec 23 12:34:41 2024 +0000" }, "message": "vscode: add gopls config file\n\nThis allows using VSCode and its derivatives to use our gopackagesdriver\nto have working code intelligence with our Bazel build. It also\noptimizes the VSCode file watcher to ignore lots of file changes in\nBazel symlinked output directories.\n\nChange-Id: I1d36a31b5c58987d42f78a80ba8831761327d884\nReviewed-on: https://review.monogon.dev/c/monogon/+/3711\nTested-by: Jenkins CI\nReviewed-by: Leopold Schabel \u003cleo@monogon.tech\u003e\n" }, { "commit": "5b997a113d011c35ecc981594c95a53d67311f83", "tree": "61b4a9db6eab7a23d58242a098b8a65670b181eb", "parents": [ "fce7c765ace0ae7058af9498e228eca30dd4695c" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Dec 19 15:10:07 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Mon Dec 23 10:13:43 2024 +0000" }, "message": "osbase/event/memory: remove Sync feature\n\nThe Sync feature was unused, and I\u0027m not sure using it anywhere is a\ngood idea. If you really do need reliable delivery of events to multiple\nconsumers, then you probably also would need some mechansim for a\nconsumer to restart at the point in the history it has last seen before\nit failed. Removing this feature simplifies the implementation and\nallows new features to be implemented.\n\nChange-Id: I94731355b7208f91256bc4610cb31a2720ab6c68\nReviewed-on: https://review.monogon.dev/c/monogon/+/3717\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "fce7c765ace0ae7058af9498e228eca30dd4695c", "tree": "4d331860270a59969d0f917a72e87e7dde74b9a6", "parents": [ "08c1c72246da4f8f18dfb7e94a1da813b094c7a4" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Dec 19 14:07:24 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Mon Dec 23 10:13:43 2024 +0000" }, "message": "osbase/supervisor: restart DONE nodes when needed\n\nNodes in DONE state are normally not restarted. However, if a sibling of\nthe node dies, then the node in DONE state and all its children are\ncanceled and need to be restarted. Previously, this did not happen, so\nthe newly added test got stuck waiting forever for the inner runnable to\nstart again.\n\nChange-Id: I8c437937d21edb822985ec6fa3f36425e12ff218\nReviewed-on: https://review.monogon.dev/c/monogon/+/3716\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "08c1c72246da4f8f18dfb7e94a1da813b094c7a4", "tree": "e8dd5a474edb79f2658b951e64b993b38c46b210", "parents": [ "65602097b716674316a318d5594a8b2b0e52d239" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Dec 19 12:03:17 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Mon Dec 23 10:13:43 2024 +0000" }, "message": "osbase/supervisor: only enter DONE state after runnable returns\n\nPreviously, a node was marked DONE as soon as it signaled DONE. If a GC\nrun happens between the time when the runnable signals DONE, and when\nthe runnable exit is processed, this causes problems. The test which I\nadded panics without the other changes:\n\npanic: could not find [inner] (root.inner) in root (NODE_STATE_NEW)\n\nIf the delay is long enough that the inner node has already restarted,\nthen this could even end up with multiple instances of the same runnable\nrunning simultaneously.\n\nI fixed this problem by only entering the DONE state after the runnable\nhas returned.\n\nChange-Id: If73b73f104c4cc204bce4374f4ba5f7e163e4a0b\nReviewed-on: https://review.monogon.dev/c/monogon/+/3715\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "65602097b716674316a318d5594a8b2b0e52d239", "tree": "750858f9b0302860f49ccde11edbd908ce58df94", "parents": [ "aa6b42a15c38d1d99301673fa76c6b0f2953a892" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Dec 19 10:37:34 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Mon Dec 23 10:13:43 2024 +0000" }, "message": "osbase/supervisor: make restart sleep cancelable\n\nPreviously, if a runnable died, and immediately after also its parent,\nthen the parent was only restarted after waiting for both the child\nand then parent restart backoff sleep. Additionally, the child was\nrestarted with an already canceled context.\n\nNow, the restart sleep can be canceled, and if canceled will directly\ngo to CANCELED state without first starting the runnable.\n\nChange-Id: Ie986db680d4df12d590881d1a7e468c741a732d9\nReviewed-on: https://review.monogon.dev/c/monogon/+/3714\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "aa6b42a15c38d1d99301673fa76c6b0f2953a892", "tree": "5b79f12996420eb010ff8bb340710f8ef4669383", "parents": [ "bdbb9c2baa94f72bf57ad1c13a2c2a5c3ff01858" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Wed Dec 18 18:03:26 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Mon Dec 23 10:13:43 2024 +0000" }, "message": "osbase/supervisor: canceled runnables are not live\n\nThis fixes a bug which caused a test to be flaky; it could get stuck\nwith the TestHarness waiting for runnables in CANCELED state to die.\n\nChange-Id: I31394e407662fe05918907650931859c6bba35fe\nReviewed-on: https://review.monogon.dev/c/monogon/+/3713\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "bdbb9c2baa94f72bf57ad1c13a2c2a5c3ff01858", "tree": "97fad3e69df9ac10117a71846320dd8748caeecd", "parents": [ "742fde7cd4861bb16b4f0655b84f587510c1e84b" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Wed Dec 18 15:14:02 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Mon Dec 23 10:13:43 2024 +0000" }, "message": "m/node/core/time: use CommandContext\n\nChange-Id: Ie98b949facf3d26c819bdf56329f5837b8e3dac7\nReviewed-on: https://review.monogon.dev/c/monogon/+/3712\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "742fde7cd4861bb16b4f0655b84f587510c1e84b", "tree": "c9a9dbf91270ac38f315636ed3acf4086431b107", "parents": [ "d7c8bbbf6638f0042a972af7efaf808b159fcbb1" ], "author": { "name": "Timon Stampfli", "email": "timon@timon.ch", "time": "Sun Dec 15 16:39:31 2024 +0100" }, "committer": { "name": "Timon Stampfli", "email": "timon@timon.ch", "time": "Fri Dec 20 15:53:04 2024 +0000" }, "message": "build/ci: add windows metroctl builds\n\nmetroctl now builds for Windows. Make sure it stays that way by\nincluding it in the CI target set.\n\nChange-Id: Id9b6a42159c6ff2400f1e8788a5ae6d2c4156e9f\nReviewed-on: https://review.monogon.dev/c/monogon/+/3698\nVouch-Run-CI: Lorenz Brun \u003clorenz@monogon.tech\u003e\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "d7c8bbbf6638f0042a972af7efaf808b159fcbb1", "tree": "4ed649b74c53ac0d1b514b2d2821556da33bef00", "parents": [ "91bcf46639db7008c7290e6f27136cd122fd1b3c" ], "author": { "name": "Timon Stampfli", "email": "timon@timon.ch", "time": "Sun Dec 15 17:26:35 2024 +0100" }, "committer": { "name": "Timon Stampfli", "email": "timon@timon.ch", "time": "Fri Dec 20 15:53:04 2024 +0000" }, "message": "m/c/metroctl: clean up ssh terminal handling\n\nDrop deprecated x/crypto/ssh/terminal package. It\u0027s replacement is\nalready used in the same package. Consistently use syscall.Stdin and\ncast it to int to be compatible with Windows. This is a no-op on\nUnix-style platforms.\n\nChange-Id: I727a51a6aae6e57e4133ab5969db77af013ba78f\nReviewed-on: https://review.monogon.dev/c/monogon/+/3699\nVouch-Run-CI: Lorenz Brun \u003clorenz@monogon.tech\u003e\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "91bcf46639db7008c7290e6f27136cd122fd1b3c", "tree": "f0edbfddd7a96ce00edb0db0bb1cf294536d9f1a", "parents": [ "007d66e6b32db8080d37738b2986871729a48e03" ], "author": { "name": "Timon Stampfli", "email": "timon@timon.ch", "time": "Sun Dec 15 16:57:05 2024 +0100" }, "committer": { "name": "Timon Stampfli", "email": "timon@timon.ch", "time": "Fri Dec 20 15:53:04 2024 +0000" }, "message": "m/node: remove non-definition dependencies\n\nThis enables usage from arbitrary platforms without including lots of\ndependencies that aren\u0027t related to functionality.\n\nChange-Id: I33e16b5396dc7216b676b294b8c1752caf3551b3\nReviewed-on: https://review.monogon.dev/c/monogon/+/3697\nTested-by: Jenkins CI\nVouch-Run-CI: Lorenz Brun \u003clorenz@monogon.tech\u003e\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "007d66e6b32db8080d37738b2986871729a48e03", "tree": "2f8e97539a365a16eab8c84a5be701c1303b07a6", "parents": [ "83f0aa47b5e9c02cc6871760fcac29b31b344c48" ], "author": { "name": "Timon Stampfli", "email": "timon@timon.ch", "time": "Sun Dec 15 16:29:04 2024 +0100" }, "committer": { "name": "Timon Stampfli", "email": "timon@timon.ch", "time": "Fri Dec 20 15:53:04 2024 +0000" }, "message": "o/blockdev: add windows implementation\n\nThis adds a relatively bare bones, but working implementation of\nblockdev for Windows. Has been manually tested on Windows 23H2.\n\nChange-Id: I7a2bf0a46e973f838506a321ad815629f62a3b35\nReviewed-on: https://review.monogon.dev/c/monogon/+/3696\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nVouch-Run-CI: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "83f0aa47b5e9c02cc6871760fcac29b31b344c48", "tree": "a3f923ef7db47733008c92d2d2a7fc0d24b32ec8", "parents": [ "b62b8e04eb6f2f6ebc54ecc397ded788a924f279" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 09 20:34:33 2024 +0100" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Dec 17 23:06:51 2024 +0000" }, "message": "osbase/build: update progress messages to contain output short_name\n\nChange-Id: Id41a05a7b8f5cec6e50d9c74ee09309d1a4b504f\nReviewed-on: https://review.monogon.dev/c/monogon/+/3691\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "b62b8e04eb6f2f6ebc54ecc397ded788a924f279", "tree": "9934baf66b686eee0609ec2ceb402450de0afee3", "parents": [ "b9701c362d602b9b51961bcff849b2eb28b65883" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Dec 16 20:18:47 2024 +0100" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Dec 16 20:24:07 2024 +0000" }, "message": "m/n/kubernetes: fix flake in TestAsFlags\n\nThis test was flaky due to Go\u0027s map iteration not being deterministic.\nSort the output to make sure we do not introduce unnecessary\nnon-determinism.\n\nFixes: #363\nChange-Id: If70486306a809b7d33bc17206600b0f750429b7d\nReviewed-on: https://review.monogon.dev/c/monogon/+/3708\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "b9701c362d602b9b51961bcff849b2eb28b65883", "tree": "536fd1e47e35ca5b7ee81b6097b619c721b6916a", "parents": [ "c441f0b4141a4908bbae142dcde929e427ad8fd0" ], "author": { "name": "Timon Stampfli", "email": "timon@timon.ch", "time": "Sun Dec 15 17:50:01 2024 +0100" }, "committer": { "name": "Timon Stampfli", "email": "timon@timon.ch", "time": "Mon Dec 16 19:34:45 2024 +0000" }, "message": "m/c/metroctl: add prototext node-params option\n\nThis adds the node-params flag to metroctl install, which allows\nadvanced users to specify customized NodeParameters in a prototext file.\nThe cluster field gets overwritten by the existing logic. As of now,\nthis is only useful for specifying network_config.\n\nChange-Id: Ieccf208177bb49635a634e484c67e99e80792bda\nReviewed-on: https://review.monogon.dev/c/monogon/+/3700\nVouch-Run-CI: Lorenz Brun \u003clorenz@monogon.tech\u003e\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "c441f0b4141a4908bbae142dcde929e427ad8fd0", "tree": "159d2da9f1b27a01dd8f22f4a2cc1c97fb7c6c12", "parents": [ "b7df5ba641a78b4ef0c926ea45bffb5cf76c8915" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Dec 04 18:02:26 2024 +0100" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Thu Dec 05 13:08:05 2024 +0000" }, "message": "gomod: switch ethtool back to upstream\n\nOur privflags patch got into the 0.2.0 relase, use that and drop our\noverride.\n\nChange-Id: I4fe83874f3ddfe66dcff006df86570ab946a30ff\nReviewed-on: https://review.monogon.dev/c/monogon/+/3681\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "b7df5ba641a78b4ef0c926ea45bffb5cf76c8915", "tree": "96f9f7a280d661de9f00cd6b17b158db49033b34", "parents": [ "d1a8b64d305c57f45416fc40b39211541113a373" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Dec 04 17:58:54 2024 +0100" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Thu Dec 05 13:08:05 2024 +0000" }, "message": "gomod: clean up indirect section\n\nWe had some direct dependencies in the indirect section, causing go mod\ntidy to create new sections every time a new dependency is added.\n\nClean this up by moving the direct dependencies into the correct\nsection, leaving us with one direct and one indirect section.\n\nChange-Id: I1b06e7b08d8d7ca2ed535400d33997b089e04cf0\nReviewed-on: https://review.monogon.dev/c/monogon/+/3680\nTested-by: Jenkins CI\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\n" }, { "commit": "d1a8b64d305c57f45416fc40b39211541113a373", "tree": "17fcd0e77576b200e75a940fb26ce2334a7a8553", "parents": [ "d77e26ee216738393a9808c95266bbcb91ca0e68" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Tue Dec 03 17:40:41 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Wed Dec 04 08:28:03 2024 +0000" }, "message": "treewide: add more ptr.To usages\n\nChange-Id: Ibf511bc012a17e39d6b7b4f3a7d9abc1304d755f\nReviewed-on: https://review.monogon.dev/c/monogon/+/3677\nTested-by: Jenkins CI\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\n" }, { "commit": "d77e26ee216738393a9808c95266bbcb91ca0e68", "tree": "8dd5dfa48c9b388684b697687be4198094ac66e3", "parents": [ "affe8fa229e3a701e060cb6bc35b9362814b5daf" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 02 18:23:10 2024 +0100" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Dec 03 14:31:57 2024 +0000" }, "message": "treewide: replace bool-to-boolptr helpers with k8s.io/utils/ptr.To\n\nChange-Id: I90419ddfe087291f41f7f2f3589263e56c15470a\nReviewed-on: https://review.monogon.dev/c/monogon/+/3675\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "affe8fa229e3a701e060cb6bc35b9362814b5daf", "tree": "63c91aa39676df52139f5c0df6023feff1ab0def", "parents": [ "e49c273818f88cc4ff2fd9fcb8bb8272b70f1beb" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Dec 02 17:53:24 2024 +0100" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Dec 03 14:31:57 2024 +0000" }, "message": "osbase/build: remove redundant platform_transition_binary rule\n\nWe have the aspect library in our dependencies anyway,\nso there is no point in copying this.\n\nChange-Id: I160a13a90dd45a0495670bd453a3aa0a16cfb771\nReviewed-on: https://review.monogon.dev/c/monogon/+/3674\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "e49c273818f88cc4ff2fd9fcb8bb8272b70f1beb", "tree": "513aa18ea3661eca62df782c1f064ee8b5c84c06", "parents": [ "2ecccae4ff62b687ec5e218349fcf8a42069dfc9" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Thu Nov 28 20:01:39 2024 +0100" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Dec 03 14:31:57 2024 +0000" }, "message": "treewide: remove duplicate test_output flag\n\nChange-Id: Ia45b46dbbdb2daad2da33094acfe047b6ad09ab3\nReviewed-on: https://review.monogon.dev/c/monogon/+/3673\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "2ecccae4ff62b687ec5e218349fcf8a42069dfc9", "tree": "c5a5914c9d3bd8fb37a5650a6b3e4881f9fc2610", "parents": [ "d58edf4e2f745427d69ecc72bfe9a9ead69d697d" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Nov 27 22:03:35 2024 +0100" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Dec 02 16:50:54 2024 +0000" }, "message": "m/node: enable user namespaces in K8s\n\nThis enables the two feature gates for user namespace support in K8s.\nWe did not previously have a passwd file which caused Go\u0027s UserLookup\nto fail with an unexpected error. Add an mostly-empty placeholder file\nto placate it.\n\nChange-Id: I71a7a6dc889a289512075a25b7e551f2cd65ffb6\nReviewed-on: https://review.monogon.dev/c/monogon/+/3665\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "d58edf4e2f745427d69ecc72bfe9a9ead69d697d", "tree": "bd9424fdb0a58cb7c78ab99d8a3b1d4ebc07c5db", "parents": [ "ff7452b586134e18af9f1362d7b96dcb64aa8d71" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Nov 27 20:38:14 2024 +0000" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Dec 02 16:50:54 2024 +0000" }, "message": "m/n/kubernetes: introduce feature gate infra\n\nThis introduces centralized infrastructure to control feature gates in K8s.\n\nIt includes a test to make sure that we do not keep outdated flags in there.\n\nChange-Id: Ife251cbd5210bc8b3757bb3829e91bcdb2e6fdfb\nReviewed-on: https://review.monogon.dev/c/monogon/+/3664\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "ff7452b586134e18af9f1362d7b96dcb64aa8d71", "tree": "7e3b9fe5c161cedf1073a086d0b6e5511b20bd98", "parents": [ "231ee041b652ab2aea6a64e0c4929fa4beb5851b" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Nov 28 13:08:55 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Nov 28 14:45:57 2024 +0000" }, "message": "m/node/kubernetes: mount PVs with noexec on the host\n\nNow that runc always replaces per-mount-point flags when bind-mounting\nvolumes inside the container, we can mount them with noexec on the host\nwithout affecting workloads. This has some security advantages, as any\nexecutables in volumes are no longer executable from the host.\n\nChange-Id: Id5a8ea8caf702fca58d300fc9e17c21e94ebaf13\nReviewed-on: https://review.monogon.dev/c/monogon/+/3660\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "231ee041b652ab2aea6a64e0c4929fa4beb5851b", "tree": "c0d279637a526450cf60004cae32aa00edba57d6", "parents": [ "690c42d8e98c4b9ad5caec1f8dc0da91f9347f12" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Nov 20 19:10:35 2024 +0000" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Thu Nov 28 11:46:04 2024 +0000" }, "message": "workspace: update K8s to 1.31.3\n\nRelevant changes:\n* Reverted DisableNodeKubeProxyVersion feature gate to off\n* Various scheduler preemption fixes and improvements\n* Lifecycle fix for pods with init containers getting stuck in creating\n when the container runtime returns an error.\n* Kubelet no longer processes terminated pods when it is restarted to\n prevent lifecycle bugs.\n* Fixed endpoint controller not reconciling endpoints after the object\n gets truncated (more than 1000 endpoints).\n* Fixes a memory leak related to tracing.\n* Fixes a crash in controller-manager\u0027s service-lb-controller loop.\n\nChange-Id: I287a532dd3781079b351cc71ca95d596ad2b59b3\nReviewed-on: https://review.monogon.dev/c/monogon/+/3626\nTested-by: Jenkins CI\nReviewed-by: Jan Schär \u003cjan@monogon.tech\u003e\n" }, { "commit": "690c42d8e98c4b9ad5caec1f8dc0da91f9347f12", "tree": "58a3e2cbd510aaa5286a82f983e26ba2d83c8553", "parents": [ "7873f46b185f846260d0119fad34e8882a48bf8d" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Nov 21 12:10:53 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Nov 28 09:58:03 2024 +0000" }, "message": "metropolis/node: validate label prefixes with our own function\n\nI think it makes sense to use our own domain validation function here\ninstead of using the function from Kubernetes. The Kubernetes one is\nless strict than ours, and actually allows names which are not valid\nDNS names, because it does not limit the length of labels to 63.\n\nAll labels which are valid according to ValidateLabelKey should also be\nvalid according to Kubernetes IsQualifiedName, and I added a test for\nthis. We need this property for synchronizing labels to Kubernetes.\n\nChange-Id: I0f96551b7d41f38b28174b7349cd8f37e6fd8f81\nReviewed-on: https://review.monogon.dev/c/monogon/+/3624\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "7873f46b185f846260d0119fad34e8882a48bf8d", "tree": "125fdf48f5b678357bb33548d5100718ce60c819", "parents": [ "73beb693ce8aed1c1caffaec2f01b2b9c65516b3" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Nov 20 20:02:31 2024 +0100" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Nov 27 19:34:17 2024 +0000" }, "message": "workspace: update runc to 1.2.2\n\nContains various fixes as well as support for transparent user\nnamespaces with K8s.\n\nChange-Id: I7441cda74ddbaa38712a1e594884f37fa87892b6\nReviewed-on: https://review.monogon.dev/c/monogon/+/3625\nReviewed-by: Jan Schär \u003cjan@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "73beb693ce8aed1c1caffaec2f01b2b9c65516b3", "tree": "378d3b779febf33b1666438b1dd003053d9fd21c", "parents": [ "be70c9247b7c8f7ab0eef4b0c7b1faaf934b8f97" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Wed Nov 27 17:47:09 2024 +0100" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Nov 27 19:34:17 2024 +0000" }, "message": "m/node/kubernetes: remove local-strict storage class\n\nIt turns out that the local-strict storage class did not have an effect\non readonly volumes, or on gVisor. And after updating runc to 1.2.0, it\nno longer has an effect anywhere. It appears that setting noexec and\nsimilar flags in the CSI server, using a storage class, is the wrong\napproach and just happened to work by accident. Instead, this should\nprobably be implemented as a Kubernetes feature to set per-mount-point\nflags on the VolumeMount.\n\nThis commit thus removes the local-strict storage class and the mount\noptions processing in the provisioner and CSI server. This will allow\nupdating runc.\n\nAdditionally, the StatefulSet end-to-end test is extended to also run\ntests with gVisor. gVisor apparently does not support block volumes.\n\nSee: https://github.com/monogon-dev/monogon/issues/361\nChange-Id: Ic2f50aa3bc9442ca1dbb9e8742d5b8fecbfc3614\nReviewed-on: https://review.monogon.dev/c/monogon/+/3658\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "be70c9247b7c8f7ab0eef4b0c7b1faaf934b8f97", "tree": "b1126b8ddaf845314329bd33249e2ec0db6940dd", "parents": [ "0ec0c53061acd57cf545440a723c1fd9817ed080" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Nov 21 11:16:03 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Nov 21 12:57:42 2024 +0000" }, "message": "m/node/kubernetes: fix attaching block PVs\n\nAttaching a block PV to a container failed with the error:\n\"failed to create device node at target path: file exists\".\nThis happened because there was already a directory at the path.\nThe directory should only be created for mounts, not for block devices.\n\nI also extended the PV end-to-end test to add a block volume, and check\nthat it can be opened as a block device and has the expected size.\n\nChange-Id: I40ca82cfcbfee1cb3196a900423f967b45790a64\nReviewed-on: https://review.monogon.dev/c/monogon/+/3623\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "0ec0c53061acd57cf545440a723c1fd9817ed080", "tree": "ac07fa1b10948234fe1add7300508a427c058325", "parents": [ "652c2ad2e499ca709523978e04b3a3dbb6df642c" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Thu Aug 29 12:39:47 2024 +0000" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Nov 20 18:40:12 2024 +0000" }, "message": "m/n/k/containerd: upgrade to v2\n\nUpgrade containerd to 2.0, migrate config and adjust all paths.\nNo new K8s features are enabled yet, this will come separately.\n\nAlso bumps gVisor to the latest version and essentially reimplements the\nshim as the API has changed a lot.\n\nA drive-by fix in clitable was necessary as the x/tools upgrade\nintroduced a new analysis pass.\n\nChange-Id: I9d25af203b94667aaac69a71eeccad2d42aa5f99\nReviewed-on: https://review.monogon.dev/c/monogon/+/3622\nTested-by: Jenkins CI\nReviewed-by: Jan Schär \u003cjan@monogon.tech\u003e\n" }, { "commit": "652c2ad2e499ca709523978e04b3a3dbb6df642c", "tree": "4a31c1797694ed53331d1a998922c3587d940d5b", "parents": [ "36f0375c9834d82016cb077142d2eaaea981d7a5" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Tue Nov 19 17:40:50 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Wed Nov 20 13:55:19 2024 +0000" }, "message": "m/node/kubernetes: fix PV mount flags and add e2e test\n\nMount flags did not work because of two problems:\n- The provisioner did not copy them from the StorageClass to the\n PersistentVolume.\n- The CSI server used \u003d instead of |\u003d when adding flags, so only one of\n the flags was added or removed.\n\nThere was an existing e2e test for PVs, however this only created the\nPVC/PV without even attaching it to a container. I extended this test to\nattach the PV and check from inside the container that it has the\nexpected mount flags and quota.\n\nThe existing e2e test also created a block PV, however attaching a block\nPV to a container was not tested and is apparently broken, so I removed\nthis test for now.\n\nChange-Id: Ie14adfafd333eab38d2b5f1b4ce8a2aa8795eae0\nReviewed-on: https://review.monogon.dev/c/monogon/+/3613\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "36f0375c9834d82016cb077142d2eaaea981d7a5", "tree": "301dacf15758e619d2c7fc0901e90ce6a5229f81", "parents": [ "1fc5eb01ef91a4c6fc936a177329b845f0bb45d7" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Tue Nov 19 17:41:05 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Wed Nov 20 13:55:19 2024 +0000" }, "message": "m/test/e2e/suites/kubernetes: increase test timeouts\n\nThese timeouts are too short and cause spurious test failures.\n\nChange-Id: I21252337a150bb5774ef031eea53ec1a366f7fbd\nReviewed-on: https://review.monogon.dev/c/monogon/+/3614\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "1fc5eb01ef91a4c6fc936a177329b845f0bb45d7", "tree": "a482438d1f4475952d52f1b11957124fd67700f2", "parents": [ "1587a80c5a13a64798b46e32ecad998dd96db906" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Nov 12 13:21:52 2024 +0100" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Nov 12 19:13:52 2024 +0000" }, "message": "treewide: use toolchain resolution to access protoc\n\nChange-Id: I7eb5cfe2619dba7b56c7952c2042313a6e8db6ac\nReviewed-on: https://review.monogon.dev/c/monogon/+/3610\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "1587a80c5a13a64798b46e32ecad998dd96db906", "tree": "6d685e516c6e125dc5ccfc46dde0f291b7824be1", "parents": [ "795951910e1c6f66efecf40e4dcc909d143999fc" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Sep 30 21:18:03 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Nov 12 19:02:38 2024 +0000" }, "message": "metropolis: add boot IDs to status\n\nThis allows for precisely determining if a kernel restart has occurred.\nUseful for making tests more accurate and relying less on sleeps.\n\nCloses: #357\nChange-Id: Ic215b5db841b29b3a3c622333a05be6c35cc6ded\nReviewed-on: https://review.monogon.dev/c/monogon/+/3477\nTested-by: Jenkins CI\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\n" }, { "commit": "795951910e1c6f66efecf40e4dcc909d143999fc", "tree": "c6b16e84340b390da57b3628f6d5ef1b71ac89c1", "parents": [ "d589b6afab60a4cfc5723c04ae83183a4487ec7e" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Mon Nov 11 14:55:56 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Tue Nov 12 16:14:55 2024 +0000" }, "message": "osbase/fat32: write file creation time\n\nAdd file creation times to directory entries. This was already partially\nimplemented, and there even was a test. But the test was conditional on\nthe kernel version, and the version comparison was broken. The test\ntimestamps were also wrong, as the last component of time.Date is\nnanoseconds, not milliseconds.\n\nChange-Id: Ic00d6bb27aced918df018c9ff3148d12ecb07753\nReviewed-on: https://review.monogon.dev/c/monogon/+/3608\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "d589b6afab60a4cfc5723c04ae83183a4487ec7e", "tree": "fd404d8cdda2f86a1298aaa85b15a3946770e597", "parents": [ "5d1bf6ab597d1ebfc4284bcb7df6648c17535464" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Mon Nov 11 14:55:38 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Tue Nov 12 16:14:55 2024 +0000" }, "message": "osbase/fat32: fix corrupted directories\n\nPreviously, after writing the contents of a file or directory, any\nremaining space in the last cluster was left as is instead of zeroed.\nFor files this may be acceptable, but directories do not have an\nexplicit size in bytes. The result is that, if the device was not zeroed\nbefore writing the FAT32, each directory contains some number of\nadditional garbage entries, with a random name, size, and initial\ncluster.\n\nThe fix is to ensure that the last cluster of a directory is filled up\nwith zeroes, following the FAT32 spec (page 24): \"initialize all bytes\nof that cluster to 0\". The change also applies to files, where zeroing\nthe remainder of the last cluster should not be necessary, but also\ndoesn\u0027t hurt. At most 32 KiB of zeroes (the maximum cluster size) will\nbe written for each file.\n\nThe fsck test is extended to initialize the file with random bytes.\nWithout the fix, this results in fsck reporting a large number of\nproblems.\n\nChange-Id: I7729e028c92bfad9b879a4256f3fa4f7af25553a\nReviewed-on: https://review.monogon.dev/c/monogon/+/3607\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "5d1bf6ab597d1ebfc4284bcb7df6648c17535464", "tree": "1f57a604f9f3e9420e8d7bde27d8ae5fed31ea98", "parents": [ "275facf8c7a5f3dc826e27e75a6122226acd5694" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Tue Nov 12 13:34:40 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Tue Nov 12 13:14:50 2024 +0000" }, "message": "build/ci: add bazel mod tidy step\n\nRunning `bazel mod tidy` on the current main branch causes a change in\nbuild/bazel/rust.MODULE.bazel. If `bazel run //:go -- mod tidy` changes\ngo.mod, it will also also run `bazel mod tidy`, which means that this\nwill eventually be noticed in a later unrelated change. But it is better\nif this is noticed and fixed immediately, so I added `bazel mod tidy` to\nthe steps run by CI. It looks like we then don\u0027t need `bazel mod deps`\nanymore.\n\nI also moved the git status step to the post always block, such that it\nis run even if one of the previous steps fails. For example,\n`bazel run //:gazelle -- update` can fail if a change to\nMODULE.bazel.lock is necessary. This should make it easier to understand\nwhy the pipeline failed, because you can then still see which files have\nchanged, and the \"Unclean working directory\" message is printed.\nI tested that this works, and that the pipeline still fails if a step\nin the always block fails.\n\nChange-Id: I05e3f1a5d9e45ba16058cc6758b0acb106458cae\nReviewed-on: https://review.monogon.dev/c/monogon/+/3609\nTested-by: Jenkins CI\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\n" }, { "commit": "275facf8c7a5f3dc826e27e75a6122226acd5694", "tree": "a4f5068b0d1888e9c81757539bc52f0dcf8e5f89", "parents": [ "a8938da203b9ecc42a61b4aa9e92b802bf0e4902" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Sep 24 00:24:29 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Nov 11 16:03:59 2024 +0000" }, "message": "m/cli/metroctl/test: rename test file to fit naming scheme\n\nChange-Id: Ia4b93203d0da27ef171272bb9cb18ac7cdac5db0\nReviewed-on: https://review.monogon.dev/c/monogon/+/3511\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "a8938da203b9ecc42a61b4aa9e92b802bf0e4902", "tree": "52c8f2971cc6ce50b9bf17a490a7defbf66e69d2", "parents": [ "9eab31ccbba4a2db416e4c5c387d22ec672ea92f" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Fri Sep 13 22:34:01 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Nov 11 16:03:55 2024 +0000" }, "message": "metropolis/node/kubernetes: add mountOptions support for PVs\n\nWe have very strict defaults on our data mount which prevents exec\u0027s and\nsuid binaries. By adding support for mountOptions on PVs we enable\nthe user to allow specific behaviour e.g. exec\u0027s on the given PV.\n\nChange-Id: I902cf3b9dafb14598cddc18c327ef3f5bcd6450b\nReviewed-on: https://review.monogon.dev/c/monogon/+/3421\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "9eab31ccbba4a2db416e4c5c387d22ec672ea92f", "tree": "92b81e90fdb13d29a12dc4ef1a253e40eddd10ff", "parents": [ "272c8301e0db375689dbc5bee6134b91cc23188d" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Wed Oct 23 12:33:36 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Wed Nov 06 16:47:06 2024 +0000" }, "message": "treewide: bump rust to 1.82.0\n\nThis also patches rules_rust to support extra_rustc_flags inside bzlmod\nextensions.\n\nChange-Id: I5da146bee580128a7ecfa6d98fa3afe482deb112\nReviewed-on: https://review.monogon.dev/c/monogon/+/3599\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "272c8301e0db375689dbc5bee6134b91cc23188d", "tree": "31c97490efe77ff7b571402f51b776af2d601b97", "parents": [ "b701df98b1706751142f29bee032321447886267" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Nov 05 05:17:44 2024 +0100" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Wed Nov 06 16:47:06 2024 +0000" }, "message": "osbase/build/mkimage: replace embedsrc abloader reference with argument\n\nrules_rust things that our abloader target itself is an exec target\nbecause it is included as embedsrc inside mkimage. To prevent this wrong\ndetection we provide it as runfile like the kernel and rootfs. This is\na preparation for updating rust to the current stable version, as it\nrequires specific overrides to work correctly with our efi toolchain.\n\nChange-Id: I78de6a15570a81d9f673702ec4e50954d604117d\nReviewed-on: https://review.monogon.dev/c/monogon/+/3598\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "b701df98b1706751142f29bee032321447886267", "tree": "27b650e77b9be17108f8e03aecfd535ec0366242", "parents": [ "78567601398f4db5a8080fd30038ff7ac6affe0f" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Thu Oct 31 14:15:33 2024 +0000" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Tue Nov 05 13:11:03 2024 +0000" }, "message": "metropolis/metroctl: implement cluster configure\n\nThis is a framework for simple ClusterConfiguration changes via\nmetroctl. We only have one mutable field for now\n(kubernetes.node_labels_to_synchronize), but more fields can be\nsupported later.\n\nThis could also be extended to support operations like \u0027add\u0027 and\n\u0027remove\u0027 for repeated fields.\n\nFinally, there could be another CLI command that would drop you into a\nprototext editor, similar to `kubectl edit xxx`. But this solves the\nsimplest usecase for now.\n\nChange-Id: I2fc588a2a2249a5c4f0cf52acb162cac9ed3d9a4\nReviewed-on: https://review.monogon.dev/c/monogon/+/3595\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "78567601398f4db5a8080fd30038ff7ac6affe0f", "tree": "757ee7c8d374317366a2535dbfb48ceaa66700f0", "parents": [ "beec27c6bdc2da730ffa2a2be6a68e1610148913" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Thu Oct 31 13:42:04 2024 +0000" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Tue Nov 05 13:11:03 2024 +0000" }, "message": "metropolis: remove stutter in ClusterConfiguration.KubernetesConfig\n\nWe already know this is a config (it lives in ClusterConfiguration), no\nneed to call that a config again.\n\nThis doesn\u0027t break any compatibility yet as field names are not (yet)\nunder a stability guarantee.\n\nChange-Id: Ib6492d1c8303cbd0620b979b8047ec9757e301c0\nReviewed-on: https://review.monogon.dev/c/monogon/+/3594\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "beec27c6bdc2da730ffa2a2be6a68e1610148913", "tree": "d9c130b03107cdcbadcc50ff850d96d0c83b8dcb", "parents": [ "2db687330d6c907bf31006730bdd161cd34cddaa" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Thu Oct 31 12:27:08 2024 +0000" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Tue Nov 05 13:11:03 2024 +0000" }, "message": "metropolis/cli/metroctl: move takeownership under cluster subcommand tree\n\nThis is in preparation for having more cluster-wide commands (same as we\nkeep \u0027node\u0027 for all node-specific commands).\n\nChange-Id: I3fa5935e1d751261bf82a80cf61e811fa36effc9\nReviewed-on: https://review.monogon.dev/c/monogon/+/3593\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "2db687330d6c907bf31006730bdd161cd34cddaa", "tree": "996bd64c955d1d383cd2bddb296d41437186a82f", "parents": [ "492434add0746eef043e5e5acc106035ccec4358" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Wed Oct 30 18:25:33 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Oct 31 14:48:23 2024 +0000" }, "message": "osbase/build/genosrelease: Add trailing newline to os-release\n\nText files should end with a newline.\n\nChange-Id: I14f06359959a41d3a49ecfb3923cb5e509020310\nReviewed-on: https://review.monogon.dev/c/monogon/+/3592\nTested-by: Jenkins CI\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\n" }, { "commit": "492434add0746eef043e5e5acc106035ccec4358", "tree": "8a6cb3b702b55e69f7e78f57a6ece0196bdf6567", "parents": [ "1f51cf42fcd4d7bb7f4b103c797c438bfa7b0098" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Oct 22 14:29:55 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Thu Oct 31 14:09:37 2024 +0000" }, "message": "treewide: remove qemu build\n\nWe aren\u0027t bundling it right now and it is fairly out of date. Let\u0027s kick\nit out until we need it, then reengineer the build.\n\nChange-Id: I5362616922d4b9d3e971868636f3792e33880f7a\nReviewed-on: https://review.monogon.dev/c/monogon/+/3568\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "1f51cf42fcd4d7bb7f4b103c797c438bfa7b0098", "tree": "8baeed7e76e82440e9217ea7055d04af0ba14435", "parents": [ "39f4f5c360e7a286bff4adaeabc52393dc28dc22" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Oct 01 17:04:28 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Thu Oct 31 14:09:31 2024 +0000" }, "message": "treewide: update rules_rust to v0.53.0\n\nThis updated our patches for rules_rust, removes a transition as it can\nbe replaced with the \"platform\" field in the rust_binary rule. This then\nallows us to correctly reference it in all targets that depend on it.\nAdditionally the -target parameter is replaced inside the llvm-efi\ntoolchain with --target\u003d.\n\nChange-Id: Ie98753e505736c9ef28ff92fa1c5aa5b3612aec3\nReviewed-on: https://review.monogon.dev/c/monogon/+/3473\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "39f4f5c360e7a286bff4adaeabc52393dc28dc22", "tree": "c81382f2408a83ab7391414738713633b8fc9608", "parents": [ "1e39914fbcecda7ec236e67f143bbefc31eee9da" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Tue Oct 29 09:41:50 2024 +0100" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Wed Oct 30 13:10:29 2024 +0000" }, "message": "metropolis: add cluster domain config and metroctl param\n\nThis adds a --cluster parameter to metroctl and a cluster domain field\nto the bootstrap configuration. It is not yet used anywhere, but later\nthe cluster domain will be used to identify the cluster.\n\nThe length of the cluster domain is limited to 80, to allow for\nconstructing subdomains. This limit could be increased later if needed,\nbut it cannot easily be decreased, so I chose a conservative value that\nshould be enough in most cases.\n\nChange-Id: I627cca8eb1d92c4b06e4dfd6b6926a013e8f33ae\nReviewed-on: https://review.monogon.dev/c/monogon/+/3508\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "1e39914fbcecda7ec236e67f143bbefc31eee9da", "tree": "806a09d23eec324d7ff131f42ddfab13cc0f98e0", "parents": [ "20498ddc40079451c83ba3708afc57d820866cb3" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Tue Oct 22 10:58:15 2024 +0000" }, "committer": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Wed Oct 30 11:42:51 2024 +0000" }, "message": "metropolis: first pass API for reconfiguring cluster\n\nThis implements management.ConfigureCluster. This API is based around\nProtobuf FieldMasks, which is a new thing in the Metropolis codebase\n(node config mutation is performed via optional fields).\n\nWhether this is the right way to do this is to be discussed.\nAlternatives considered are:\n\n1. Always insert a full new config, providing the old one as a base. The\n downside of that is the potential conflicts that will spring up the\n moment we have systems regularly mutate independent parts of the\n config. Additionally, this might lead to some odd behaviour when\n dealing with clients that don\u0027t have support for newer versions of\n the config proto.\n2. Use optional fields, like in Node role code. However, this has the\n downside of duplicating protos (one for the config state, one for the\n mutation request). Plus, protobuf optionals are still somewhat\n unusual.\n3. Provide individual requests for mutating fields (like with Node\n labels). This also results in a lot of boilerplate code.\n4. Something akin to JSON Patch, but for protobufs, which doesn\u0027t seem\n to exist.\n\nChange-Id: I42e5eabd42076e947f4bc8399b843e0e1fd48548\nReviewed-on: https://review.monogon.dev/c/monogon/+/3591\nTested-by: Jenkins CI\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\n" }, { "commit": "20498ddc40079451c83ba3708afc57d820866cb3", "tree": "08849c86dba480fd56e4252a302ee804d8ac7fae", "parents": [ "e99638e3c7a2f1a604d49c47cc7a2685bfff8c5e" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Sep 30 17:07:08 2024 +0000" }, "committer": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Oct 28 14:22:49 2024 +0000" }, "message": "metropolis/test/launch: synchronize test node labels to kubernetes\n\nChange-Id: I57361503debb1d25873a72be18e947d2d1f11511\nReviewed-on: https://review.monogon.dev/c/monogon/+/3470\nReviewed-by: Jan Schär \u003cjan@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "e99638e3c7a2f1a604d49c47cc7a2685bfff8c5e", "tree": "636c243a58100c971cc3e224abf2c54324aad00a", "parents": [ "9579be5e09b6293edc78d3142b0c67a24afda93c" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Sep 30 17:06:44 2024 +0000" }, "committer": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Oct 28 14:22:49 2024 +0000" }, "message": "metropolis/node/kubernetes: synchronize metropolis node labels to kubernetes\n\nThis extends the labelmaker to manage Kubernetes node labels mirrored\nfrom Metropolis node labels.\n\nNote that currently there is no way to edit a ClusterConfiguration at\ncluster runtime, but this will come in a future CL.\n\nChange-Id: If7dbc3796085a8b85c1b5b2a181bcb1cee3d1db4\nReviewed-on: https://review.monogon.dev/c/monogon/+/3469\nReviewed-by: Jan Schär \u003cjan@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "9579be5e09b6293edc78d3142b0c67a24afda93c", "tree": "52fd3bd699099ff599eca2d7c52febad6b5b41c4", "parents": [ "dd2b80fa4eb22931702aae5d849c178a4930e101" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Sep 30 17:01:04 2024 +0000" }, "committer": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Oct 28 14:22:49 2024 +0000" }, "message": "metropolis/proto/common: add node label synchronization rules\n\nThis paves the way for a mechanism to synchronize Metropolis node labels\nto Kubernetes node labels. This is just the API/Protobuf part.\n\nChange-Id: Ia6f5dd91190d46495714ea56aa359c48e6a068d7\nReviewed-on: https://review.monogon.dev/c/monogon/+/3468\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "dd2b80fa4eb22931702aae5d849c178a4930e101", "tree": "56d10f07a4dec157756a62caa7e534a270591921", "parents": [ "6d1ff36763f1d48cf8620afd17321a06d2fbe228" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Tue Sep 24 13:06:27 2024 +0000" }, "committer": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Oct 28 14:22:49 2024 +0000" }, "message": "metropolis: support prefixes in node labels\n\nThis brings Metropolis node label semantics to be the same as Kubernetes\nlabels.\n\nChange-Id: I33c321432ec01abf978bb8dfbb3cef90f75a38eb\nReviewed-on: https://review.monogon.dev/c/monogon/+/3467\nTested-by: Jenkins CI\nReviewed-by: Jan Schär \u003cjan@monogon.tech\u003e\n" }, { "commit": "6d1ff36763f1d48cf8620afd17321a06d2fbe228", "tree": "e0f48b5b138f51579de1ce2662e1b3a39acec6d3", "parents": [ "677de978403a58cd219e77b312b647927bd560ac" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Sep 30 15:15:31 2024 +0000" }, "committer": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Oct 28 14:22:49 2024 +0000" }, "message": "metropolis/node/kubernetes: update labels based on node roles\n\nThis implements the labelmaker, a reconciling loop running on Kubernetes\ncontroller nodes which updates Kubernetes node labels based on cluster\ndata.\n\nCurrently it only updates role labels based on cluster roles, but this\ncan be extended in the future to also replicate Metropolis node labels\ninto Kubernetes node labels.\n\nChange-Id: I9c5ba92bb46f064aa03836720d4a80adc6061ab9\nReviewed-on: https://review.monogon.dev/c/monogon/+/3464\nTested-by: Jenkins CI\nReviewed-by: Jan Schär \u003cjan@monogon.tech\u003e\n" }, { "commit": "677de978403a58cd219e77b312b647927bd560ac", "tree": "68c55ccd2b6680d8016cf066cb1f13feb2b60113", "parents": [ "5d0f63442a3ec8998d7ed9eb3a3b04ad127a88db" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Wed Sep 25 05:30:04 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Wed Oct 16 12:09:54 2024 +0000" }, "message": "treewide: bump go to 1.23.1\n\nAlso removing two checks which are useless for us as they validate usage\nof cockroach internal type usage.\n\nChange-Id: Ia0f920baa4dfd3068e6669f1ca5363f227ce67de\nReviewed-on: https://review.monogon.dev/c/monogon/+/3452\nReviewed-by: Serge Bazanski \u003cserge@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "5d0f63442a3ec8998d7ed9eb3a3b04ad127a88db", "tree": "f1b1f37cefe775c27cb7c3c8cfddb9d1e3697914", "parents": [ "8eebee7e32f60095137dab785a3cc3f97c85d03d" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Wed Sep 25 03:35:00 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Wed Oct 16 12:09:54 2024 +0000" }, "message": "treewide: bump bazel dependencies\n\nChange-Id: Iee54b3e74e914120ab8d644382afe40a3647ef4b\nReviewed-on: https://review.monogon.dev/c/monogon/+/3451\nTested-by: Jenkins CI\nReviewed-by: Serge Bazanski \u003cserge@monogon.tech\u003e\n" }, { "commit": "8eebee7e32f60095137dab785a3cc3f97c85d03d", "tree": "f5bd0bca5b537a56071f54b60dbad6d34b9fb4a7", "parents": [ "0bc92a087ee0eb279ab29c3aba5d127b4202a2ea" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Sep 26 10:33:48 2024 +0200" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Tue Oct 15 07:48:46 2024 +0000" }, "message": "m/n/c/curator: add ID field to nodes in etcd\n\nThis makes the ID independent of the public key for nodes stored in\netcd. This is needed to eventually allow node key rotation.\n\nWe could just extract the ID from the key without adding an ID field.\nBut the consistency check between key and value has already caught a bug\nonce, so it seems worth keeping.\n\nChange-Id: I7ba5904d37d54e93ad6dc7d4b6f0cfac19bc730d\nReviewed-on: https://review.monogon.dev/c/monogon/+/3475\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "0bc92a087ee0eb279ab29c3aba5d127b4202a2ea", "tree": "9c481ad86d6324cdd6bdfff4a55af4d4b4689f3c", "parents": [ "61b97a375aee98f58c13c13be672b442aecc8440" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Oct 01 22:53:08 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Thu Oct 10 15:55:35 2024 +0000" }, "message": "treewide: bump rules_oci to v2.0.0\n\nChange-Id: Idbeb3a3b7645c5b6f774eb43d218ca0bc79dccc1\nReviewed-on: https://review.monogon.dev/c/monogon/+/3474\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "61b97a375aee98f58c13c13be672b442aecc8440", "tree": "75e76cee9a7b32a31650f06f8b3c775d598016a4", "parents": [ "0b4fb8c4987b6ce0c8d33d9b643e0bcee5bfabe5" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Wed Oct 02 13:30:33 2024 +0200" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Wed Oct 09 12:23:11 2024 +0000" }, "message": "m/node/core/localstorage: fix EFI directory name\n\nThe EFI directory is called EFI, not ESP. The ESPEFIDirectory is not\nused anywhere, so this typo did not have any effect.\n\nChange-Id: I38cd44ee06cb5f210acbd4a608e499b0372c2633\nReviewed-on: https://review.monogon.dev/c/monogon/+/3476\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "0b4fb8c4987b6ce0c8d33d9b643e0bcee5bfabe5", "tree": "e812207ac2a8573b3058e6a213a4810d8f65866c", "parents": [ "0de92127b6b247c4e2a816a84a1ff045a603ca73" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Wed Sep 18 17:34:23 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Oct 08 21:00:43 2024 +0000" }, "message": "metropolis/cli/metroctl: refactor to use RunE instead of log.Fatal\n\nChange-Id: Id5ca65980816e1715a8f08afcdf712292117012a\nReviewed-on: https://review.monogon.dev/c/monogon/+/3441\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "0de92127b6b247c4e2a816a84a1ff045a603ca73", "tree": "438e9666ad45536f95bd85c54556ee58d428adff", "parents": [ "a512b0e03b370ea86b383212863d81b6d9677f3a" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Sep 24 06:37:30 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Oct 08 21:00:39 2024 +0000" }, "message": "treewide: replace static mentions of repo name with dynamic ones\n\nThis does change the CONFIG_QEMU_FIRMWAREPATH to \"/nonexistant\", but\nsince we are using OVMF everywhere anyway this should behave the same.\n\nCloses monogon-dev/monogon#343\n\nChange-Id: I865e5ba5e62579c3ff1f31a25e46cbcb78dba688\nReviewed-on: https://review.monogon.dev/c/monogon/+/3450\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "a512b0e03b370ea86b383212863d81b6d9677f3a", "tree": "373bb9ff7fd8e63928205c28d8480c33bfd8f2aa", "parents": [ "5fb8a3fc41a1c59636adaf55c6495c1a671ef7ad" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Sat Sep 28 03:57:43 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Oct 08 13:55:15 2024 +0000" }, "message": "metropolis/vm: delete vm package\n\nThis only contains the CRD right now and is not used/implemented\nanywhere. This is also the only usage of kube-code-gen which breaks with\nthe newest version of rules_go. Since we don\u0027t plan to use this\nin the near future this also gets deleted, because even if we need it\nagain, we can just revert this change.\n\nChange-Id: I29ab75541957fce6a7dd8414c0df3cfdf90f8ec3\nReviewed-on: https://review.monogon.dev/c/monogon/+/3465\nTested-by: Jenkins CI\nReviewed-by: Serge Bazanski \u003cserge@monogon.tech\u003e\n" }, { "commit": "5fb8a3fc41a1c59636adaf55c6495c1a671ef7ad", "tree": "102e0e764764bce2ff2e0f375500b2ef0f236ac5", "parents": [ "d5538b52d7a8739f7123458c10973be36b27b9ff" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Sep 30 17:04:20 2024 +0000" }, "committer": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Sep 30 18:18:32 2024 +0000" }, "message": "metropolis/curator: log warning if cluster configuration cannot be loaded\n\nChange-Id: I4ad520d7545fe88d0db85a223f3a2e6d51e05136\nReviewed-on: https://review.monogon.dev/c/monogon/+/3466\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "d5538b52d7a8739f7123458c10973be36b27b9ff", "tree": "fd718931af36798650ddbf1a8978a94220994e82", "parents": [ "1dcede9600e4c1584da4fbe89128970ea9532860" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Wed Sep 25 13:16:49 2024 +0200" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Sep 26 11:44:09 2024 +0000" }, "message": "m/n/c/consensus: fix startup after removing a cluster node\n\nThe consensus service was waiting for all initial peers to be DNS\nresolvable before starting etcd. However, the list of initial peers is\nnever updated. If an etcd member is removed from the cluster, it is no\nlonger resolvable, but may still be contained in initial peer lists. The\nconsensus service then fails to start, as it is blocked forever waiting\nfor the removed peer to become resolvable.\n\nThe wait for resolvability was added in c1cb37ce9c43 with this\nexplanation:\n\n\u003e It also makes the consensus service wait for DNS resolvability before\n\u003e attempting to join an existing cluster, which makes etcd startup much\n\u003e cleaner (as etcd will itself crash if it cannot immediately resolve\n\u003e its ExistingPeers in startup).\n\nThis does not appear to be needed anymore. I did not observe etcd\ncrashes after removing the wait for resolvability.\n\nI extended the e2e test to test this scenario. After removing the\nconsensus role, it also deletes the node and reboots the remaining\nnodes. I moved these tests to the ha_cold suite, because with encryption\nenabled, we currently cannot reboot a node in a 2-node cluster.\n\nChange-Id: If811c79ea127550fa9ca750014272fa885767c77\nReviewed-on: https://review.monogon.dev/c/monogon/+/3454\nTested-by: Jenkins CI\nReviewed-by: Serge Bazanski \u003cserge@monogon.tech\u003e\n" }, { "commit": "1dcede9600e4c1584da4fbe89128970ea9532860", "tree": "1561bfcc59a2663ee432339bf4d815e399571a9a", "parents": [ "39d9c24f7167eb853aed0e1865ef8b187adf5bba" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Wed Sep 25 13:07:11 2024 +0200" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Sep 26 11:44:09 2024 +0000" }, "message": "m/test/launch/cli/launch-cluster: add allow-reboot flag\n\nWith the `-allow-reboot` flag set, it is possible to reboot a node with\n`metroctl node reboot`.\n\nChange-Id: I5c16f40d231803e31452a3b3c524be209d501526\nReviewed-on: https://review.monogon.dev/c/monogon/+/3453\nReviewed-by: Serge Bazanski \u003cserge@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "39d9c24f7167eb853aed0e1865ef8b187adf5bba", "tree": "dc8229e272c2f78eac56bdf4fde135809444f255", "parents": [ "8d82f8d261b14b73385ba66e44279c53bb9fef13" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Tue Sep 24 13:49:55 2024 +0200" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Sep 26 11:44:09 2024 +0000" }, "message": "metropolis: reduce usage of identity.NodeID\n\nEventually, we want to be able to rotate node keypairs. To allow this,\nthe node ID needs to become independent of the public key. This change\nis a refactoring which starts this work by reducing the usage of\nidentity.NodeID, the function which derives a node ID from a public key.\n\nChange-Id: I5231ed0a7be37c23327fec93481b00c74374af07\nReviewed-on: https://review.monogon.dev/c/monogon/+/3445\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "8d82f8d261b14b73385ba66e44279c53bb9fef13", "tree": "f7332695f7546dd34ac043ac4ce4d6dcc6547dce", "parents": [ "ad8982fcd78a3408c024d9031fa611b48dd86304" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Wed Sep 18 11:22:46 2024 +0200" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Sep 26 11:44:09 2024 +0000" }, "message": "m/n/c/curator: maintain consistency between roles and etcd members\n\nWhen updating the consensus role, both etcd membership and the role need\nto be updated. It is possible that the etcd membership change is applied\nbut the role update fails, resulting in an inconsistency. This change\nadds a background process which cleans up this inconsistency by updating\nroles to match etcd membership.\n\nThis is partially based on previous work by Serge Bazanski, where this\nbackground sync was performed in the opposite direction: etcd membership\nis removed if the role is missing. Here, I instead update the role based\non etcd membership. This has the benefit that we finish partially\napplied management operations, instead of fighting them.\n\nCo-authored-by: Serge Bazanski \u003cserge@monogon.tech\u003e\nChange-Id: I8871b068d1d20c65bcbea5289eafe54676906819\nReviewed-on: https://review.monogon.dev/c/monogon/+/3438\nTested-by: Jenkins CI\nReviewed-by: Serge Bazanski \u003cserge@monogon.tech\u003e\n" }, { "commit": "ad8982fcd78a3408c024d9031fa611b48dd86304", "tree": "b6a7f84b0d7d8e1d4531883eac22dab990c6f1c7", "parents": [ "fc6e1cf11d0d96fac1e8d52b5787b207f8b1fd9f" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Tue Sep 17 13:56:34 2024 +0200" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Sep 26 11:44:09 2024 +0000" }, "message": "m/node/core: remove etcd membership before removing consensus role\n\nWhen removing the consensus role, we also need to remove etcd\nmembership. It is safer to remove membership first, and then the role,\nbecause otherwise, the etcd cluster is in a degraded state during the\ntime where etcd on the node has been stopped, but the node is still\ncounted as a voting member by etcd.\n\nIf the membership is removed, but then removing the role fails, the\ncluster ends up in an inconsistent state. If the affected node was the\ncurator or etcd leader, that will almost certainly happen. In this case,\nthe request can just be retried until it succeeds, and then the cluster\nstate is consistent again between etcd membership and roles.\n\nChange-Id: I1ab526470a4201e76817e8ca0a597996fb903d1f\nReviewed-on: https://review.monogon.dev/c/monogon/+/3437\nTested-by: Jenkins CI\nReviewed-by: Serge Bazanski \u003cserge@monogon.tech\u003e\n" }, { "commit": "fc6e1cf11d0d96fac1e8d52b5787b207f8b1fd9f", "tree": "85ee090a2aa8cfe6bdc64e9116066350f8d12605", "parents": [ "e551d31d39e7cf13e7adf2f66941e6989a708c60" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Wed Sep 18 17:34:07 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Sep 23 15:44:49 2024 +0000" }, "message": "metropolis/cli/metroctl: don\u0027t print usage for runtime errors\n\nChange-Id: I91fdca45e874ea1ec9112df556a9a7392fd45bfd\nFixes: https://github.com/monogon-dev/monogon/issues/352\nReviewed-on: https://review.monogon.dev/c/monogon/+/3440\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "e551d31d39e7cf13e7adf2f66941e6989a708c60", "tree": "ac04077adea875cf28f8c52f427c72c780b383ae", "parents": [ "5f1a7de2dfb5db1884fcb677a0bd38daf6dd3c97" ], "author": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Tue Sep 17 12:38:54 2024 +0200" }, "committer": { "name": "Jan Schär", "email": "jan@monogon.tech", "time": "Thu Sep 19 15:59:14 2024 +0000" }, "message": "third_party/go: patch etcd to fix panic in IsLocalMemberLearner\n\nAfter removing etcd membership from a node, etcd panics in\nIsLocalMemberLearner. Because we currently run etcd embedded in the core\nprocess, this causes the entire machine to reboot.\n\nThe panic happens in the Status RPC handler, which we call to obtain the\ncurrent etcd leader. Stack trace:\n\ngo.etcd.io/etcd/server/v3/etcdserver/api/membership.(*RaftCluster).IsLocalMemberLearner(0xc000879380)\n\texternal/gazelle~~go_deps~io_etcd_go_etcd_server_v3/etcdserver/api/membership/cluster.go:859 +0x26d\ngo.etcd.io/etcd/server/v3/etcdserver.(*EtcdServer).IsLearner(0xc001020c18?)\n\texternal/gazelle~~go_deps~io_etcd_go_etcd_server_v3/etcdserver/server.go:2813 +0x1a\ngo.etcd.io/etcd/server/v3/etcdserver/api/v3rpc.(*maintenanceServer).Status(0xc001020bb0, {0xc001d38638?, 0xc001d38638?}, 0xc001d38690?)\n\texternal/gazelle~~go_deps~io_etcd_go_etcd_server_v3/etcdserver/api/v3rpc/maintenance.go:228 +0x137\ngo.etcd.io/etcd/server/v3/etcdserver/api/v3rpc.(*authMaintenanceServer).Status(0x4a5c340?, {0x3122cf8?, 0xc0026f2f00?}, 0x2d0ebd8?)\n\texternal/gazelle~~go_deps~io_etcd_go_etcd_server_v3/etcdserver/api/v3rpc/maintenance.go:306 +0x25\n\nA node that is no longer a member is also not a learner anymore, so just\nreturn false instead of panicking.\n\nChange-Id: I6fa3d710836852772d5c2bf05dbd63a3d68b1a8b\nReviewed-on: https://review.monogon.dev/c/monogon/+/3436\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "5f1a7de2dfb5db1884fcb677a0bd38daf6dd3c97", "tree": "fd52bf35b4b2e6b5c51f56d62424c9d0820ef537", "parents": [ "e337e938ae8e08dffa3a01045571188413ce70ff" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Thu Sep 19 02:00:14 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Thu Sep 19 12:06:50 2024 +0000" }, "message": "treewide: fix %v in cases where we should use %w\n\nWe should always use %w when using fmt.Errorf as you can use error.Is to\ncompare the underlying error. When printing an error the use of %w is\nwrong and should be replaced with %v.\n\nChange-Id: I741111bd91dcee4099144d2ecaffa879fdbb34a2\nReviewed-on: https://review.monogon.dev/c/monogon/+/2993\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "e337e938ae8e08dffa3a01045571188413ce70ff", "tree": "f82fa1f5722c3eae99506510056fb6a5ce736309", "parents": [ "7a1b27df41a9729dd9669cdaabd6864afc5e85b7" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Sun Sep 15 20:14:39 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Wed Sep 18 22:27:59 2024 +0000" }, "message": "m/n/k/containerd: set device ownership based on security context\n\nWhen a user deploys a pod with a kvm device it is owned by root. By\nsetting device_ownership_from_security_context to true, containerd\nwill chown these devices to the uid/gid set in the securityContext.\nFor more informations see\nhttps://kubernetes.io/blog/2021/11/09/non-root-containers-and-devices/\n\nChange-Id: I1a0285dfc560c3c662d5e2eb8e37e68d87408b83\nReviewed-on: https://review.monogon.dev/c/monogon/+/3428\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "7a1b27df41a9729dd9669cdaabd6864afc5e85b7", "tree": "fe32eee491bd556f3379ac2fba779d56314f045c", "parents": [ "d0e39cbad2e12cbb23caa7d36ccbd4821285ac51" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Thu Feb 22 23:54:58 2024 +0100" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Sep 18 19:39:51 2024 +0000" }, "message": "metropolis/cli/metroctl: implement install ssh\n\nThis implements another way of installing metropolis via ssh. It does\nthis by uploading the files to the target machine and then doing a kexec\ninto the install environment. If it fails at any point it will print the\nerror and reboot.\n\nChange-Id: I1ac6538896709c386b053a84903fa04940c1f012\nReviewed-on: https://review.monogon.dev/c/monogon/+/2079\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "d0e39cbad2e12cbb23caa7d36ccbd4821285ac51", "tree": "7d233926774a9af8ca065a7dacc533f22f6a5a85", "parents": [ "e4b1d20497b31ce639b6d8d8fb7079ea49686144" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Mon Sep 16 16:14:00 2024 +0200" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Sep 18 19:39:51 2024 +0000" }, "message": "go/net/ssh: allow multiple AuthMethods\n\nChange-Id: Icdd8c12bc959f87271db4abba3219edd0ea1b7ca\nReviewed-on: https://review.monogon.dev/c/monogon/+/3434\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "e4b1d20497b31ce639b6d8d8fb7079ea49686144", "tree": "4f143be7218443701acf6fb1830762a4a749c34d", "parents": [ "c5d28e49c53c7dd0b1f88160514342a1e1b98958" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Tue Sep 17 23:44:46 2024 +0200" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Sep 18 18:31:28 2024 +0000" }, "message": "m/node/core: fix nodeparams dependent on network\n\nThe GCP nodeparams gathering strategy depends on network availability.\nWith the introduction of static network configuration that got added to\nNodeParameters which meant that they needed to be there before the\nnetwork could be initialized. This dependency loop stalls bootup on GCP\nforever.\n\nFix it by splitting up NodeParameter gathering into a local and a\nnon-local phase. In setups where metadata is gathered via network\nautomated network configuration is generally always available to break\nthis dependency loop. Thus we can start networking after the local phase\nhas finished and run the non-local (i.e. networked) phase later.\n\nChange-Id: I661b9b474f67f2289f427327efa4c3eaa19393e7\nFixes: https://github.com/monogon-dev/monogon/issues/353\nReviewed-on: https://review.monogon.dev/c/monogon/+/3439\nReviewed-by: Leopold Schabel \u003cleo@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "c5d28e49c53c7dd0b1f88160514342a1e1b98958", "tree": "6e3c355798c5b8113fb44ed790ba6e9d58ad82a7", "parents": [ "ca8d951b683a3f0c64da7f61d4f74567d50623ac" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Tue Sep 17 20:38:31 2024 +0200" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Tue Sep 17 21:36:22 2024 +0000" }, "message": "go/net/ssh: improve throughput by more than 100x\n\nUse concurrent writes (this is safe for our usage) and use\nReadFromWithConcurrency to extract maximum concurrency from the process.\n\nIncrease the window to 1024 pending reqeuests for a bandwidth-delay\nproduct of 32MiB, sufficient for ~1Gbps at 300ms RTT.\n\nIn practice this improves upload performance from ~13Mbit/s on 15ms RTT\nup to over 1.5Gbit/s, now being limited by the weak network\nimplementation on the small GCP instance I tested this on. Maximum\nperformance with a bigger instance and \u003c1ms RTT was over 5Gbps, becoming\ndisk-limited.\n\nChange-Id: I2f08d2c815fd15794f497ed9dc7a7f8a2e351992\nReviewed-on: https://review.monogon.dev/c/monogon/+/3435\nTested-by: Jenkins CI\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\n" }, { "commit": "ca8d951b683a3f0c64da7f61d4f74567d50623ac", "tree": "8e8f7af5a5902c0807d77d6774dfd8b426510624", "parents": [ "04aa3df595521dab1fe8fb12b716d2826a37105f" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Thu Sep 12 14:20:57 2024 +0200" }, "committer": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Sep 16 16:29:59 2024 +0000" }, "message": "metropolis/resolver: use logging.Leveled\n\nThis moves the resover client library to use logging.Leveled instead of\nan ad-hoc logger interface.\n\nBy now having multiple level of logs, and by defaulting metroctl to show\nerrors and warnings, this should fix #302.\n\nChange-Id: I7cae1cf1be377ec824ad46ea1da1b23b46e01903\nReviewed-on: https://review.monogon.dev/c/monogon/+/3432\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "04aa3df595521dab1fe8fb12b716d2826a37105f", "tree": "3e7dc34d941c09c7693ea8016108311de8e65f1b", "parents": [ "f9c824951877cf295d7fd031073c2a3c0e09c804" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Thu Sep 12 14:07:25 2024 +0200" }, "committer": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Sep 16 16:29:59 2024 +0000" }, "message": "go/logging: add backends\n\nThese are (for now unused) backends for the common Leveled interface.\n\nChange-Id: Ic45e7ea26ad0143500489b05ac6bb20aa7fdbcfb\nReviewed-on: https://review.monogon.dev/c/monogon/+/3431\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "f9c824951877cf295d7fd031073c2a3c0e09c804", "tree": "0b8f0405bb2094d90b7008a594ba00ff31f0144a", "parents": [ "c7a332be2ac37af8d3e27029e6300c8884988ab5" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Sep 16 16:50:39 2024 +0200" }, "committer": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Sep 16 15:28:27 2024 +0000" }, "message": "third_party/linux: build at 16 threads, request 16 CPUs\n\nThis makes sure we don\u0027t overwhelm build machines with a bunch of\nparallel kernel builds. That should in turn deflake tests which are\nresource-dependent (like E2E tests).\n\nChange-Id: If2758f328d7b2c71c8f405246c6ac41091b4a350\nReviewed-on: https://review.monogon.dev/c/monogon/+/3433\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "c7a332be2ac37af8d3e27029e6300c8884988ab5", "tree": "dc834fee3e970b047b65e329d8f278cd5a91eeca", "parents": [ "3c5d0635f855f16780792a6be311f71b4d59f20b" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Thu Sep 12 17:58:42 2024 +0200" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Sep 16 15:13:57 2024 +0000" }, "message": "third_party/linux: disable fallback tunnel devs\n\nIntroduces a patch for a kernel config option to control the default\nbehavior for fallback tunnel devices in the kernel.\nWe want to fully disable them but this can otherwise only be done by\npassing kernel commandline arguments which would need to be specified in\na lot of places and can easily be missed.\n\nThis will get rid of things like the sit0 interface in the host and\ncontainer namespaces.\n\nChange-Id: I2c03c5aa50bc64b527b72c3c022d3e9f2111dd84\nReviewed-on: https://review.monogon.dev/c/monogon/+/3430\nTested-by: Jenkins CI\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\n" }, { "commit": "3c5d0635f855f16780792a6be311f71b4d59f20b", "tree": "4a48292bf17a874f2d627901ee4f7e9145c5b040", "parents": [ "a036c4e792e4b497c512991291b0cc18bc12b5e3" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Thu Sep 12 10:49:12 2024 +0000" }, "committer": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Mon Sep 16 14:03:22 2024 +0000" }, "message": "osbase/logtree.LeveledLogger -\u003e go/logging.Leveled\n\nThis factors out the common leveled logger interface out of the logtree.\nWe want to use the same interface outside of logtree/supervisor usage\nwithin the resolver code, which will be exposed to clients.\n\nChange-Id: I299e76d91e8cefddf8f36f1e58432418c4694df2\nReviewed-on: https://review.monogon.dev/c/monogon/+/3411\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "a036c4e792e4b497c512991291b0cc18bc12b5e3", "tree": "d759d4504d042a62577a1b57a9093a4ab97f0f67", "parents": [ "96e014e23888e09c12a8f0dd78ac13a1b319751d" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Tue Sep 10 19:11:57 2024 +0200" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Mon Sep 16 12:21:19 2024 +0000" }, "message": "m/n/c/mgmt: move and factor out reboot code\n\nMake sure we do not have two copies of mostly temporary reboot code\nsitting around and put it in a sensible place.\n\nChange-Id: I293a699dbfc3cfe23378485c512d8769b2859ab8\nReviewed-on: https://review.monogon.dev/c/monogon/+/3396\nReviewed-by: Serge Bazanski \u003cserge@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "96e014e23888e09c12a8f0dd78ac13a1b319751d", "tree": "40504bb1497ed5f3624d299b6b241c2a92cc3af2", "parents": [ "58321127dc8e4c44cfb0eee372a313b277857a2e" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Sep 10 02:26:13 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Thu Sep 12 23:31:41 2024 +0000" }, "message": "metropolis/installer: migrate to bringup package\n\nChange-Id: Ib6215bc51fdef45476198eceffbd0fd1fd362f1b\nReviewed-on: https://review.monogon.dev/c/monogon/+/3393\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\nTested-by: Jenkins CI\n" }, { "commit": "58321127dc8e4c44cfb0eee372a313b277857a2e", "tree": "2cc12cfba716bd5b2501603602a299907dc66394", "parents": [ "154e6d90cd52d48d274e0fe14f070342a6c5b2b8" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Sep 10 02:26:03 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Thu Sep 12 13:56:56 2024 +0000" }, "message": "cloud/agent: migrate to bringup package\n\nChange-Id: Ibe89c37cae058395723c9ed5ad5914d62a9af891\nReviewed-on: https://review.monogon.dev/c/monogon/+/3392\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "154e6d90cd52d48d274e0fe14f070342a6c5b2b8", "tree": "ebbbb10d975d6f84cfa56bd45db904379d2ddb84", "parents": [ "437d62480bbd4e34b443e7380071bc0c41c5a948" ], "author": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Wed Sep 11 17:26:31 2024 +0200" }, "committer": { "name": "Serge Bazanski", "email": "serge@monogon.tech", "time": "Thu Sep 12 09:23:42 2024 +0000" }, "message": "metropolis: prevent printk console pollution\n\nThis implements two separate approaches to limit printk pollution of the\nnew tconsole:\n\n 1. Sets the minimum printk level to EMERG. Everything lower than this\n level will not get blasted to tty0.\n 2. Jut in case something does a spurious EMERG printk (or something\n just writes to tty0), we redraw the console. This makes it\n self-healing.\n\nChange-Id: I69370ebf6c3cb3cacc8b6ea1ad3703e758bbf50c\nReviewed-on: https://review.monogon.dev/c/monogon/+/3398\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "437d62480bbd4e34b443e7380071bc0c41c5a948", "tree": "276b2cf12b559106623bde38e11eec3287bebd1a", "parents": [ "c752ec63559ecd9b486cc9df2cdb37366ceda427" ], "author": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Tue Sep 10 02:26:27 2024 +0200" }, "committer": { "name": "Tim Windelschmidt", "email": "tim@monogon.tech", "time": "Wed Sep 11 20:20:14 2024 +0000" }, "message": "m/n/c/u/e2e/testos: migrate to bringup package\n\nChange-Id: I661e4240f5fc3a40acac38250212ab892ac121c1\nReviewed-on: https://review.monogon.dev/c/monogon/+/3394\nTested-by: Jenkins CI\nReviewed-by: Lorenz Brun \u003clorenz@monogon.tech\u003e\n" }, { "commit": "c752ec63559ecd9b486cc9df2cdb37366ceda427", "tree": "a6d0e7ee37fbc883d05e43a0fac41f419ea3238b", "parents": [ "2545895a68d016eb71436e8b8ddd7e14e1861e99" ], "author": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Sep 11 16:36:43 2024 +0000" }, "committer": { "name": "Lorenz Brun", "email": "lorenz@monogon.tech", "time": "Wed Sep 11 17:13:41 2024 +0000" }, "message": "WORKSPACE: update intel_ucode to 20240910\n\nFixes a bunch of security issues and gets us to the latest version\nagain.\n\nChange-Id: I5d6f3a4a6e1bbdd4d768479d6f5c7e37c243c9d5\nReviewed-on: https://review.monogon.dev/c/monogon/+/3407\nReviewed-by: Tim Windelschmidt \u003ctim@monogon.tech\u003e\nTested-by: Jenkins CI\n" } ], "next": "2545895a68d016eb71436e8b8ddd7e14e1861e99" }